Sie sind auf Seite 1von 11
127 CORRELATIONS BETWEEN ESP AND MEMORY J. Blackmore University of Utrecht I have discussed elsewhere the importance of attempts to integrate ESP into psychology (Blackmore, 1980). One of the first questions co ask may be whether ESP resenblés other cognitive processes such as nenory and perception. There are several ways in which this may be investigated. They include 1. The study of errors made in ESP, 2. Correlations between ESP and memory or other cognitive skills, 3: Studies of the effects of varying target material, 4. The use of the recall situation as a pei-conducive state and 5. The study of the effects of associative habits on ESP. Sone of these approaches have already been discussed. This paper will consider only one of these, that is the correlations between ESP and memory. Modern research on ESP and menory began after the publication of Roll's "memory theory of ESP" (Roll, 1966). This was based on the idea chat “the ESP response consists of the percipient's own menory traces and that the effect of the external (ESP) stimilus is to activate nenory traces rather than supply new ideas or images". Roll drew Several predictions from this theory anong then that “If menory traces are vehicles for psi impressions, we expect a person with good recall ability to perform well in ESP and, conversely, a good ESP subject to perform well tn memory tasks." (Roll 1966, p.510). This Prediction is relatively easy to test by correlating performance on ESP and menory tasks and this has been done. But before considering the research that has been carried out I wish to consider what conclusions might be drawn if such a correlation were found. Firstly Roll predicted a positive correlation between recall and ESP fon the basis of his ‘nemory theory of ESP', but I do not believe this necessarily follows from that theory. If the role of the ESP stimulus is to revive, and in so doing to select, the appropriate menory trace then the major task of recall ts carried out paranormally and there {s 0 particular reason to suppose that the same person will be good at nornal recall and this other paranormal process. On a two-process theory of enory we may argue that Koll's theory involves paranormal retrieval but nornal recognition and one might then predict a posteive correlation between ESP and neasures of recogaition, bur not recall. While this is not the only possible interpretation of Roll"s theory £¢ does indicate that a positive or negative correlation between ESP and recall scores is not necessarily evidence either for or against it. There are several other models which relate ESP and nemory. For example Carington's 'association theory of telepathy" suggests that associations between ideas and images (psychons) tend to persist independently of the person who initiated then and may be utilised by others so leading to the occurrence of telepathy. Although this theory has many drawbacks and can only accouat for telepathy, not clairvoyance, it treats menory and telepathy as equivalent processes and $0 would, I belteve, predict a positive correlation betueen thes The same nay be said of Price's "Psychic ether hypothesis’ (Price, 1939), Marshall's physical theory of ESP and newory (Marshall, 1960) and vhat Roll (1966) calls the pet-trace theory of menory, that is that the brain acts as a kind of token object in memory, allowing for retrieval of memories stored as psi-traces. Since in all these theories ESP and menory are seen as the sane process I believe they ight all predict a positive correlation between ESP and nenory. Rao, Morrison and Davis (1977) argued for a negative correlation on the basis of Bergson's idea that the brain's selective processes act aS a protection against recalling and experiencing too michs If ES? is Seen as the result of errors escaping cortical surveillance then better ESP would occur with a less efficient menory. But even this ts not a simple deduction. Bergson (1911) attributed a selective process to the nervous system, suggesting that in renenbering st selects the desired information from all that available. If all infornation 1s Potentially available then randon escapes through the surveillance would not lead to ESP but to confusion. For correct ESP sone sort of selection is still needed. Therefore a better selective process nay be able to select ESP information better as well as menory. This would lead to the opposite prediction to that nade by Rao and his associates. ESP AND MEMORY 129 If a consistent correlation between ESP and memory were found it would not be obvious which theortes of ESP would be supported or refuted. Indeed I do not think there is any theory of ESP so clearly formulated that {t would allow definite conclusions to be dravns Nonetheless, such a finding could be useful. It would indicate vhether there was asinilarity between the processes involved in ESP and henory, or in the factors which affect them, and this could provide a starts * He may now consider the existing evidence for such a correlations Feather (1965; 1967) was the first to report a correlation between ESP and wemory. She gave subjects tuo runs of clairvoyance, @ memory test for a list of 25 ES? symbols and then two nore runs of Clatrvoyance, and found a positive correlation between recall and ESP scores. This study was conplicated by the fact that the memory test vas very hard, with sone subjects scoring below chance on {ts Rao et al (1977) suggested that ESP nay have operated in the menory test as well so complicating any conclusions drava, since the correlation could have been due to ESP operating in both tests rather than to any relationship between ESP and nenory~ Many subsequent studies have tested the correlation between ESP and senory but mostly only as a secondary or even post hoc analysis. For example Stanford (1970) tested several hypotheses in one experinents Subjects Listened to a story followed by a set of multiple choice questions. The ‘correct’ answer to these questions was randonly determined. For some the answer was given in the story, for sone it vas inplied and for others it was not nentioned. According to the - negative response bias hypothesis one would expect that higher ESP Scores would be obtained for those questions which were included in the story (when the ‘correct' answer was counter-story) than for those implied or not mentioned, because menory for the story would produce a bias against then. In addition it was expected that subjects with « better memory would show a greater effect. 30 subjects completed the ESP test, a test of incidental learning for the details of a room and various other tasks, not relevant heres ‘The results confirmed the negative response bias Hypothesis. When the answer was specified in the’ story other answers were more often correct (that is by ESP) than when they were not specified. Also, when the answer had been given, there were few counter-story responses when the ‘correct’ answer was the sane as that given, but more when it was different. Figh menory subjects obtained significant scores on all counter-story Tesponses, while low menory subjects scored in the EUROPEAN JOURNAL OF PARAPSYCHOLOGY predicted direction but not significantly. Of most importance here Though £8 that on the crucial counter-story responses the high nenory subjects scored higher than the low menory subjects, but when no answer was given or implied in the story there was no difference. This led Stanford to conclude that menory, per se, {s not related to ESP perfornance, but the effect is indirect in terms of its relationship to response bias. Thus although Stanford found a positive relationship between ESP and memory he attributed it to an indirect effect of memory In a simpler experinent Peterson (1972) gave 28 subjects both memory and clairvoyance tasks using two different kinds of ESP synbol but Found no significant relationship between ESP and menory scores. Kanthanan{ and Rao (1975), developed 2 aethod for studying ESP and nenoty which has been nore widely adopted than any other. They argued that previous studies had peasured ESP and menory consecutively. They wanted to study their interaction trial by trial, bringing the two as Close together as possible. The original method consisted of giving subjects a menory task and incorporating an ESP test into the recall phase. Their hypothesis that ESP would occur differently on fecali-correet and recall-arong trials was tested, but will not be discussed here. I have previously criticised the rationale behind this method (Blacknore, 1980). ore relevant here is that an ESP and a recall score was Obtained for each subject. Kanthamani and Rao pooled data from 62 subjects and 4 series, correlated these scores and obtained a significant positive correlation of 0.284 (They give an associated t value of 2.876, pé.0l, 2-tailed. My calculations for their value of r give ¢=2.294, p=0:025, 2-tailed. This still shows a sigsLficant positive correlation). towever, this finding should be treated with caution since the analysis was conducted post hoc, and the data fron four dissimilar experiments were pooled. Parker (1976) tested 40 subjects’ digit span in an innediate memory task consisting of progressive learning of lists of 3 to 9 digits. An incorporated ESP task consisted of placing the digit answer on one of two lines. Although overall ESP scores were at chance a significant negative correlation between ESP and memory scores vas obtained (r=-0.347, £92+28, pé-05). In a second similar series a non-significant positive correlation (r=0.107) was obtained. Rao, Morrison and Davis (1977) carried out two sets of experinents in which subjects menorised lists of patred associates, of a trigran and a word. Sone ostensibly recall trials were actually ESP trials and the results of each type were correlated. In the first set of seven ESP AND MEMORY it experinents ESP scores vere close to chance as were the correlations with nenory scores and there was no apparent consistency in the direction of the correlations. In the second set, comprising seven further experinents, data for 118 subjects were pooled and (post hoc) 4 negative correlation between ESP and menory scores obtained (22-018, t=1.99, p<.05, 2-tailed). Finally in a simtlar experinent Rao (1978) tested both ESP and memory for 25 pairs of a trigram and a word--93 subjects were tested in three groups and a snall positive correlation between menory and BSP scores obtained (r=0.18, t=1.70). In addition the data fell into two groups, the first being only computer scored and the second hand scored first. It was the data fron the second group which contributed nost to the correlation. Rao discussed this difference in terns of a possible checker effect or differences in the subject-experinenter relationship. He also pointed out "Thus, the research has cone full circle: no significant correlation in the exploratory study, significant negative corcelation in the replication study, and now 4 Suggestive positive correlation for overall results and a strong positive correlation for group 2.” (Rao 1978, p.176). Whether or not one considers the correlation ‘suggestive’, these results clearly show that there ts no simple answer to the question of whether menory and ESP ability are correlated. Indeed the results of Rao and his colleagues confirn the confusing pattern seen in the previous results, that is, a positive correlation found by Feather (1965; 1987), no'correlation found by Peterson (1972) and a negative correlation found by Parker (1976). What is one to make of these results? One could conclude simply that there is no correlation between the two abilities and that the saall correlations obtained were spurious. Alternatively the correlations could be valid, but the results obscured by other variables. Perhaps nost inportant of these variables to consider is the type of nenory task used. Feather's very difficult task of learning 25 ESP syobols in 15-20 seconds bears Little resenblance to learning paired associates or testing digit span. These task differences could be the cause of the varied results, although there ust then be sone other explanation for the fact that Rao and his colleagues obtained very varied results using the same nenory task. But ignoring the latter problen for the monent, what task variables night be important? One possibtlity ts task difficulty. If ESP can be compared with Temenbering it 1s like very difficult renenbering. Therefore a harder